How to Set Up Context.dev MCP in Cursor and Claude

Context.dev’s hosted MCP server enables Cursor, Claude, and Claude Code to retrieve current web, company, and structured data through OAuth-authenticated Streamable HTTP connections at `https://mcp.context.dev/mcp`, without requiring a local package, scraper, or pasted API key. The guide updates earlier API examples to current endpoints, emphasizing `POST /v1/web/scrape` for selectable page formats and `POST /v1/web/answers` for research, while explaining MCP tools such as web-search, web-scrape-markdown, web-crawl, web-extract, sitemap discovery, document parsing, batch jobs, and brand retrieval. It provides setup and verification instructions for supported clients, recommends selecting the smallest tool suited to the task, and demonstrates searching, scraping JavaScript-rendered pages, extracting schema-validated data, and conducting bounded crawls. It also distinguishes MCP’s agent-oriented workflow from direct REST API or SDK use in application code, outlines troubleshooting for authentication, tool availability, caching, timeouts, credit limits, and stale results, and advises secure credential handling, constrained request scopes, source and request-ID tracking, and caution when treating retrieved web content as external input.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.